Reflections on the Value of M&E
One of the important points that Bernie’s paper brings to light is the fact that there are important issues in science that demand a rational, non-arbitrary resolution, but which science itself cannot resolve.
Now most academics recognize that science informs, but does not settle, moral issues. Consequently, most of Bernie’s comments were aimed at showing that they are important issues for scientists to resolve in a rational way. In my comments, I want to focus on a perhaps less socially pressing, but nevertheless important, point of contact between philosophy and science. And I want to draw a conclusion from these reflections, namely, that the scope of a rational worldview far outstrips anything that can be produced by purely empirical investigation.
We all know that observation can support general correlations between events. As a simple example: suppose we observe events of type C regularly followed by events of type E. In this case we can inductively support the conditional claim that “If a C-type event occurs, then an E-type event will occur (with some probability or other).” Such statements, in Noam Chomsky’s terminology, are descriptively adequate; that is, they give us an accurate description of how future observations will depend on past or current observations.
Now, while descriptive adequacy is an important desideratum of scientific investigation, it is by no means all we want. Science, we want to say, is not merely descriptive, but explanatory. But this transition from establishing the truth of simple descriptive claims to establishing claims that can support an explanatory framework, is a big one (at least as big as the step from “is” to “ought”); it is a step that already involves us inextricably in non-empirical philosophical issues.
Consider. Arguably a conditional like “If a C-type event occurs, then an E-type event will occur” is explanatory only if it is “law-like”. But what is a law? There are no settled answers here, but almost everyone would agree that laws involve evaluating the truth some counterfactual claims; claims about what would happen, if such and such were to happen. Now counterfactual claims involve us in considerations of unrealized possibilities, so they are not in any obvious or straightforward way settled by observation.
The point I want to drive home is this: It is not that scientists need to consult philosophers to decide which theory we ought to accept; rather, it is that scientists surreptitiously employ philosophical, non-empirical methods when deciding on theories. On the assumption that an explicit understanding of the grounds for our various beliefs is better than vague sense of those grounds (or, worse, an outright misunderstanding of them), it would seem to follow that scientists stand to gain substantially from engaging the philosophical literature on these issues (causation, modality, explanation).
But I believe that philosophy has even more substantial things to offer. We have just been looking at the step between simple observable regularities and laws. But laws themselves come in stronger and weaker forms. Some laws, for instance, are mere causal regularities (speaking picturesquely, true in some nearby worlds, but false in others) while other laws are logical necessities (true in all worlds). And this difference in modal strength between laws is crucial for deciding some very important questions. Consider, for instance, the mind-body problem. The mind-body problem is a question about how mental properties or mental states are related to physical properties or physical states.
One possibility is that mental properties are identical to physical or functional properties of some sort. This is the view of physicalists. Another possibility is that mental properties are distinct from physical or functional properties. This is dualism. How are we to decide between these two theories? Notice that an identity claim commits us to a very strong correlation between mental and physical properties: a logical correlation. If “two” things are identical, then they are necessarily identical. Consequently, even if cognitive neuroscience were to establish that there is a law-like connection between mental properties and physical or functional properties, this would not settle the mind-body problem. What we really need to settle is the strength of the connection between the two sets of properties. And this, I submit, is something that no amount of scientific investigation could ever possibly settle. Our only possible recourse is to some type of a priori philosophical reasoning.
If that is right, then certain questions which we believe have nonarbitrary, defensible answers cannot be settled by scientific investigation alone. A complete view of the world (a theory of everything) is not provided by physics or science alone, but only in tandem with philosophical theory.